Relationship Between Global Solar Radiation and Sunshine Hours for Calabar, Port Harcourt and Enugu, Nigeria

Full Length Research Paper

Relationship between global solar radiation and sunshine hours for Calabar, Port Harcourt and Enugu, Nigeria

Augustine C.* and Nnabuchi M. N.

Department of Industrial Physics, Ebonyi State University, Abakaliki, Nigeria.

Accepted 25 March, 2009

The monthly mean daily data for global solar radiation and sunshine hours for a period of seventeen years (1999 - 2007) for Calabar, Enugu and Port Harcourt respectively have been used to develop a number of regression equations. The values of the global solar radiation estimated by the models and the measured solar radiation were tested using the mean bias error (MBE), root mean square error (RMSE) and mean percentage error (MPE) statistical techniques. The values of the correlation coefficient (R) and coefficient of determination (R2) were also determined for each equation. The equations with the highest values of R, R2 and least values of MBE, RMSE and MPE for Calabar, Port 2 2 ΗΡ n ≈ n ’ ΗΡ n ≈ n ’ Harcourt and Enugu are = −0.499 + 4.294 − 4.769∆ ÷ , = 0.177 +1.048 −1.070∆ ÷ ΗΟ Ν «Ν ◊ ΗΟ Ν «Ν ◊ 2 ΗΡ n ≈ n ’ ΗΡ n and = 0.177 +1.048 −1.070∆ ÷ respectively. Where is clearness index and is ΗΟ Ν «Ν ◊ ΗΟ Ν fraction of sunshine hour. The equations could be employed in estimating global solar radiation of location that has the same geographical location information as Calabar, Port Harcourt and Enugu.

Key words: Sunshine hours, clearness index, global solar radiation.

INTRODUCTION

Energy is the motive force behind the sustained technolo- At next 40 years, the oil reserves would be depleted to gical development of any nation and Nigeria is blessed the point where it would be economically unviable to con- with reasonably high quantities of various energy resour- tinue exploration. The environmental consequences of ces. These include the non-renewables such as crude oil, harnessing these non–renewable energy sources are as- natural gas, coal and uranium and the renewables such suming alarming proportions. Oil extraction has created as biomass, solar, wind and hydro energy. At present, the so much ecological problems that the inhabitants of the dominant energy source used in Nigeria is oil and its Niger Delta region of Nigeria are beginning to think that derivatives, accounting for over 75% of the total energy coming from the oil rich area is a curse rather than a consumption, except in the rural areas where biomass in blessing (UNIDO, 2003). the form of fuel wood dominates (UNIDO, 2003). The awareness of the limited availability of these re- Energy trends indicates that world oil production will resources and their associated environmental problems is ach peak and start a long downward slide some day making it imperative that we shift emphasis to the rene- when the fossil fuel and gas would have exhausted. wable energy resources. Renewable energy technology is capable of alleviating the already over stretched ecos- stem and capable of supplying the energy necessary for rapid development, especially the rural areas by encoura- *Corresponding author. E-mail: [email protected]. ging the establishment of cottage industries and by 07038939264, 08035852721 stemming the rural urban drift. One of the most viable

Augustine and Nnabuchi 183

options particularly in Nigeria is the abundant solar at the top of the atmosphere and H is the amount of the energy falling on the surface of the earth. radiation eventually available at the ground surface after In any solar energy conversion system, the knowledge scattering, reflection, and absorption in the atmosphere. of global solar radiation is extremely important for the op- The ratio, H/HO, is a possible measure of the transpa- timal design and the prediction of the system perfor- rency of the atmosphere to solar radiation. Thus, the ratio mance. is used to define the coefficient of transmission or the A global study of the world distribution of global solar transmittance of the atmosphere (Babatunde et al., radiation requires knowledge of the radiation data in va- 1990). rious countries and for the purpose of world wide mar- A number of correlations involving global solar radiation keting, the designers and manufacturers of solar equip- and sunshine hours for different locations in Nigeria have ment will need to know the mean global solar radiation been studied by different workers. For example, (Sambo, available in different and specific regions (Ibrahim, 1985). 1985) developed a correlation with solar radiation using Obviously, the best way of knowing the amount of glo- sunshine hours for Kano with the regression coefficients bal solar radiation at a site is to install pyranometer at a = 0.413 and b = 0.241 for all the months between 1980- many locations in the given region and look after their 1984, (Arinze and Obi, 1983) developed a correlation day to day maintenance and recording, which is a very with solar radiation using sunshine hours in Northern costly exercise. The alternative approach is to correlate Nigeria with regression coefficients a = 0.2 and b = 0.74, the global solar radiation with the meteorological parame- (Burari et at., 2001) developed a model for estimation of ters at the place where the data is collected. The resul- global solar radiation in Bauchi with regression coeffi- tant correlation may then be used for locations of similar cients a =0.24 and b = 0.46. Other workers (Ojosu, 1984; meteorological characteristics. Fagbenle, 1990; Folayan, 1983; Adebiyi, 1988; Turton, Solar radiation passing through the atmosphere to the 1987; Bamiro, 1983) developed theoretical and empirical ground surface is known to be depleted through scat- correlations of broad applicability to provide solar data for tering, reflection, and absorption by the atmospheric con- systems design in most Nigeria cities. We observed that stituents like air molecules, aerosols, water vapour, the regression coefficients are not universal but depends ozone and clouds. The reflection of solar radiation is on the climatic conditions. This work aims at investigating mainly by clouds and this plays an overriding part in redu- suite of models that can use to relate global solar radia- cing the energy density of the solar radiation reaching the tion and sunshine hours for Calabar, Port Harcourt and surface of the earth (Exell, 2000). Enugu. The encounter of solar radiation particularly with clouds lead to the variation in intensity of sunshine and the num- METHODOLOGY ber of sunshine hours at the ground surface: The variation however is not due only to the clouds but also to the The monthly mean daily data for sunshine hours for Calabar, Port angle of incidence of the sun’s rays with the ground Harcourt and Enugu were collected from the archives of the Nige- surface and its azimuth (Babatunde, 1988). These in turn, rian Meteorological Agency, Federal Ministry of Aviation, Oshodi, are due to the rotation of the earth round the sun and the and Lagos. The global solar radiation data were obtained from renewable energy for rural Industrialization and development in Nige- inclination of its axle with the plane of its orbit round the ria. The data obtained, covered a period of seventeen years (1991- sun. The result is the variation in the number of hours of 2007) for Calabar (latitude 4.97oN, longitude 88.35oE and altitude sunshine and its intensity on the earth’s surface. The va- 6.14 m above sea level), Port Harcourt (latitude 4.850N, longitude riation is from latitude to latitude. Thus, a solar radiation 7.02oN and 19.55 m above sea level) and Enugu (latitude 7.55oN, o measurement parameter is obtained and defined as the longitude 6.47 E and altitude141.50 m above sea level). ratio of the actual number of hours of sunshine received at a site to the number possible in the day that is the Data analysis length of the day. The ratio is known as fraction of sunshine hours n . It is found to vary daily and seasonally The monthly averages data processed in preparation for the N correlations are presented in Tables 1, 2 and 3 for Calabar, Port (Igbal, 1979; Shears et al., 1981). Harcourt and Enugu respectively. To develop the models, the global solar radiation data measured The radiation arriving on the ground directly in line with in (Kwhm-2day-1) were converted to (MJm-2day-1) using a conversion the solar disk is called direct or beam radiation. A portion factor of 3.6 proposed by Iqbal (1983). of the scattered and reflected radiation goes back to The first correlation proposed for estimating the monthly mean space and a portion reaches the ground from the sky he- -2 -1 daily global solar radiation on a horizontal surface Η (MJm day ) misphere as diffuse radiation. The sum total of direct and using the sunshine duration data is due. Angstrom (1924) and diffuse radiation is called global radiation (Igbal, 1983). Prescott (1940) have put the Angstrom correlation in more The amount of total (global) solar radiation, H, eventually convenient form as: obtained on the ground surface is found to be to be a fraction of the extraterrestrial radiation, HO, at the top of ΗΜ n the atmosphere of the site. Extraterrestrial radiation is the =a+b (1) maximum amount of solar radiation available to the earth ΗΟ Ν

184 Int. J. Phys. Sci.

Table 1. Meteorological data and global solar radiation for Calabar.

Month Η Μ n Ν n ΗΜ ΗΟ Κ Τ =

Η Ο (hours) (hours) Ν (MJm-2day-1) (MJm-2day-1) Jan 3.889 11.75 0.3309 14.0004 34.28 0.4084 Feb 4.546 11.84 0.3839 16.3656 36.06 0.4538 Mar 4.292 11.97 0.3586 15.4512 37.52 0.4118 Apr 4.544 12.11 0.3752 16.3584 37.48 0.4365 May 4.206 12.23 0.3439 15.1416 36.24 0.4178 Jun 3.636 11.72 0.3102 13.0896 35.13 0.3726 Jul 3.233 11.74 0.2754 11.6388 35.61 0.3268 Aug 3.415 11.89 0.2872 12.2940 37.05 0.3318 Sept 3.747 11.98 0.3128 13.4892 37.26 0.3620 Oct 3.925 11.87 0.3307 14.1300 36.18 0.3905 Nov 3.983 11.76 0.3387 14.3388 34.38 0.4171 Dec 3.684 11.71 0.3146 13.2624 33.19 0.3996

Table 2. Meteorological data and global solar radiation for Port Harcourt.

Month n Ν n ΗΜ ΗΟ ΗΜ Τ -2 -1 -2 -1 Κ = (hours) (hours) Ν (MJm day ) (MJm day ) ΗΟ Jan 4.72 11.75 0.4017 14.40 34.25 0.4204 Feb 4.81 11.84 0.4063 16.26 35.50 0.4580 Mar 4.32 11.94 0.3609 15.16 36.52 0.4151 Apr 4.70 12.11 0.3881 16.68 37.28 0.4474 May 4.99 12.22 0.4083 15.16 36.54 0.4149 Jun 4.12 12.28 0.3355 13.96 35.13 0.3974 Jul 2.56 12.25 0.2089 12.99 35.71 0.3638 Aug 2.41 12.11 0.1990 12.52 37.15 0.3370 Sept 3.33 11.98 0.2779 14.02 37.56 0.3733 Oct 4.22 11.87 0.3555 14.29 35.18 0.4062 Nov 5.48 11.77 0.4656 14.00 34.48 0.4060 Dec 5.72 11.72 0.4881 14.37 32.39 0.4437

Where a and b are regression constants, ΗΜ is the measured 24 ≈π ’ monthly mean daily global radiation, n is the monthly mean daily ΗΟ= ΙscEΟ∆ Wssinφsinδ+cosφcosδsinws÷ bright sunshine hours, Ν is the maximum possible monthly mean π «180 ◊ (2) n daily sunshine hours or the day length, is the fraction of sun-

Ν Where Ιsc is the solar constant, EΟ is the eccentricity correction shine hours, ΗΟ is the monthly mean extraterrestrial solar factor, φ is the latitude, δ is the solar declination and ws is the radiation on horizontal surface, given by Iqbal (1983) as follows: hour angle.

Augustine and Nnabuchi 185

Table 3. Meteorological data and global solar radiation for Enugu.

Month n Ν n ΗΜ ΗΟ ΗΜ Τ -2 -1 -2 -1 Κ = (hours) (hours) Ν (MJm day ) (MJm day ) ΗΟ Jan 6.35 11.58 0.5484 14.25 35.82 0.3978 Feb 6.55 11.74 0.5579 15.65 37.01 0.4229 Mar 5.99 11.95 0.5013 14.77 37.54 0.3934 Apr 6.45 12.18 0.5296 14.27 36.44 0.4916 May 6.49 12.37 0.5247 14.85 34.41 0.4316 Jun 5.35 12.46 0.4294 13.61 33.15 0.4106 Jul 3.86 12.40 0.3113 11.65 34.85 0.3343 Aug 3.78 12.24 0.3088 10.80 35.47 0.3044 Sept 4.45 12.02 0.3702 12.26 36.95 0.3318 Oct 5.94 11.79 0.5038 15.18 37.73 0.4023 Nov 7.50 11.61 0.6459 16.51 35.83 0.4608 Dec 6.93 11.54 0.6005 15.42 35.22 0.4378

The accuracy of the estimated values was tested by calculating the solar radiation and clearness index for Calabar, Port Har- mean bias error (MBE), root mean square error (RMSE), and mean court and Enugu are presented in Tables 1, 2 and 3 res- percentage error (MPE). The expressions for the MBE (MJm-2day- 1 -2 -1 pectively. Tables 4, 5 and 6 contains summaries of va- ), RMSE (MJm day ), and MPE (%) are stated by El – Sebaii et al., (2005) as follows: rious linear regression analyses, obtained from the appli- cation of equation (1) to the monthly mean values for the » ÿ variable under study for Calabar, PortHarcourt and Enu- ΜΡΕ = ƒ(Ηi,cal − Ηi,meas )⁄/ n (3) g u r e s p e c t i v ely. It is clear that the correlation coefficient R, coefficient of determination R2, MBE (MJ/m2/day), RM- 2 1 SE (MJ/m /day) and MPE (%) vary from one variable to » 2 ÿ2 another variable. RΜSΕ = Ηi,cal − Ηi,meas / n (4) …ƒ( ) ⁄Ÿ

» ÿ Linear correlation ≈ Ηi,meas − Ηi,cal ’ ΜΡΕ = …ƒ∆ Χ100÷Ÿ/ n (5) … « Ηi,meas ◊⁄Ÿ The correlation coefficient of 0.951 existing between the clearness index and monthly mean daily fraction of sunshine indicates that there is a high positive correlation be- Where Ηi,cal and Ηi,meas is the ith calculated (predicted) and tween the measured monthly mean daily fraction of sun- measured values and n is the total number of observations. Iqbal shine hours and the monthly mean daily clearness index. (1983), Halouani (1993), Almorox (2005) and Che et al. (2007) Also, the value of coefficient of determination of 0.903 im- have recommended that a zero value for MBE is ideal and a low RMSE is desirable. The RMSE test provides information on the plies 90.3% of clearness index can be accounted using short- term performance of the studied model as it allows a term by fraction of sunshine. The model for Calabar is given by term comparism of the actual deviation between the calculated values and the measured values. The MPE test gives long term performance of the examined regression equations, a positive MPE ΗΡ n values provide the averages amount of overestimation in the =0.018+1.139 (6) calculated values, while the negative values gives underestimation. Ηo Ν A low value of MPE is desirable by Akpabio et al. (2002).

The correlation coefficient of 0.852 existing between the RESULTS AND DISCUSSION clearness index and monthly mean daily fraction of sunshine indicates that there is a high positive correlation be- The values for sunshine hours, day length, and fraction of tween the measured monthly mean daily fraction of sun- sunshine hours, global solar radiation, extraterrestrial shine hours and the monthly mean daily clearness

186 Int. J. Phys. Sci.

Table 4. Shows regression equations and statistical indicators for Calabar.

Equations R R2 MBE RMSE MPE 0.951 0.903 0.035 0.734 0.50 ΗΡ n = 0.018 +1.139 Ηo Ν 2 0.961 0.923 0.015 0.235 0.36 ΗΡ n ≈ n ’ = −0.499 + 4.294 − 4.769∆ ÷ ΗΟ Ν «Ν ◊

Table 5. Shows regression equations and statistical indicators for Port Harcourt.

Equations R R2 MBE RMSE MPE 0.852 0.726 0.018 0.538 0.54 ΗΡ n = 0.2946 + 0.3059 ΗΟ Ν 2 0.877 0.769 0.011 0.356 0.31 ΗΡ n ≈ n ’ = 0.177 +1.048 −1.070∆ ÷ ΗΟ Ν «Ν ◊

Table 6. Shows regression equations and statistical indicators for Enugu.

Equations R R2 MBE RMSE MPE 0.857 0.735 0.308 0.692 2.20 ΗΡ n = 0.191+ 0.433 ΗΟ Ν 2 0.871 0.758 0.023 0.586 0.42 ΗΡ n ≈ n ’ = 0.177 +1.048 −1.070∆ ÷ ΗΟ Ν «Ν ◊

index. Also, the value of coefficient of determination of Quadratic correlation 0.726 implies 72.6% of clearness index can be accounted using fraction of sunshine. The model for Port Harcourt is The correlation coefficient of 0.857 exists between the given by clearness index, monthly mean daily fraction of sunshine hours, also coefficient of determination of 0.923 implies that 92.3% of clearness index can be accounted using ΗΡ n = 0.2946 + 0.3059 (7) f r a c t io n o f s u n s h ine . The model for Calabar is given by ΗΟ Ν 2 ≈ ’ Η Ρ n n (9) The correlation coefficient of 0.857 existing between the = −0.499 + 4.294 − 4.769 ∆ ÷ Η Ο Ν « Ν ◊ clearness index and monthly mean daily fraction of sun-

shine indicates that there is a high positive correlation be- The correlation coefficient of 0.857 exists between the tween the measured monthly mean daily fraction of sun- clearness index, monthly mean daily fraction of sunshine shine hours and the monthly mean daily clearness index. hours, also coefficient of determination of 0.769 implies Also, the value of coefficient of determination of 0.735 im- that 76.9% of clearness index can be accounted using plies 73.5% of clearness index can be accounted using fraction of sunshine. The model for Port Harcourt is given fraction of sunshine. The model for Enugu is given by by

≈ ’2 ΗΡ n ΗΡ n n = 0.191+ 0.433 (8) = 0 . 1 7 7 + 1 . 0 4 8 − 1 . 0 70∆ ÷ (10) ΗΟ Ν ΗΟ Ν «Ν ◊

Augustine and Nnabuchi 187

18 Equation 6 20 16 18 Equation 7

n n o

o 16

i 14 i t t a a ) ) i

i 14 y

12 y d d a a a a r d 12 r d

10 Measured SR / r Measured SR r 2 2 a a

l 10 l m m / o / 8 Predicted SR o Predicted SR J s J s

l l M M ( a

6 ( a

b 6 b o o l 4 l 4 G G 2 2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Month of the year Month of the year

18 Equation 8 20 16 Equation 9 18

n n o i 14 o 16 t i t a ) a i )

i 14 y 12 d y d a a a a r d 12 d r

/ / 10 Measured SR r

2 Measured SR r 2 a a l 10 l m m / o

8 Predicted SR / o Predicted SR J s J s

l l M M ( a 6 ( a

b 6 b o o l 4 l 4 G G 2 2 0 0 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Month of the year Month of the year

18 Equation 10 18 Equation 11 )

16 y 16 a n d / o

i 14

2 14 t a ) m i / y 12 d 12 J a a d r M /

10 Measured SR (

r 10 Measured SR 2 a n l m o / i o 8 Predicted SR 8 Predicted SR

t J s a

i l M (

a 6 d 6 a b r o

l 4 r 4 a G l

2 o 2 S 0 0 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 Month of the year Month of the year

Figure 1. (Eqn.6-11) Comparism of the measured and predicted values of correlation equations.

The correlation coefficient of 0.857 exists between the For better analysis the developed correlations will be con- clearness index, monthly mean daily fraction of sunshine sidered that have high values of correlation coefficient R hours, also coefficient of determination of 0.758 implies and coefficient of correlation R2: Equations (6), (7), (8), that 75.8% of clearness index can be accounted using (9), (10), and (11). Equations (9), (10) and (11) have the fraction of sunshine. The model for Enugu is given by highest values of correlation coefficient and coefficient of determination, while the others have low values of corre- 2 lation coefficient and coefficient of determination. Further- ΗΡ n ≈ n’ more, there is a remarkable agreement between the ob- =0.032+1.162 −0.788∆ ÷ served and the predicted values of global solar radiation (11) f o r s e v e n t e e n y e a r s f r o m o u r c o r r e l a t ions (Figure 1). Ο Η Ν «Ν◊ From Tables 4, 5 and 6, based on the highest values of

188 Int. J. Phys. Sci.

correlation coefficient, coefficient of determination and Akpabio LE, Etuk SE (2002). Relationship between solar radiation and least values of MBE, RMSE and MPE, equations (9), (10) sunshine duration for Onne Nigeria.,Turkish J. Phys. 27: 161 - 167. Almorox J, Benito M, Hontoria C (2005). Estimating of Monthly and (11) are the best regression equations suitable for Angstrom – Prescott equation coefficients from measured daily data estimating global solar radiation in Calabar, Port Harcourt in Toledo, Spain. Renewable Energy J., 30: 931 - 936. and Enugu respectively. Angstrom AS (1924). Solar and Terrestrial radiation meteorological The values of the regression coefficients obtained for society 50: 121- 126. Arinze EA, Obi SE (1983). Solar energy availability and prediction in Calabar, Port Harcourt and Enugu were found to be Northern Nigeria. Nig. J. Solar Energy 3: 3 -10. different. These differences suggest that regression co- Babatunde EB, Aro TO (1990). Characteristics Variation of total solar efficients associated with meteorological data changes radiation at llorin, Nigeria. Nig. J. Sol. Energy 9: 157 - 173. with latitude and atmospheric conditions. Babatunde EB (1988). Solar radiation modeling for a tropical station, llorin, Nigeria. Ph.D thesis pp. 32-34.

Bamiro OA (1983). Empirical relations for the determination of solar in

Ibadan, Nigeria. Sol. Energy 31(1): 85 - 94. Conclusion Burari FW, Sambo AS, Mshelia ED (2001). Estimation of Global Solar Radiation in Bauchi. Nig. J. Renewable Energy 9: 34 - 36. The monthly mean daily global solar radiation and frac- Che HZ, Shi GY, Zhang XY, Zhao JQ, Li Y (2007). Analysis of sky tion of sunshine hours have been employed in this study condition using 40 years records of solar radiation data in China. to develop several correlation equations using SPSS Theoretical and applied climatology 89: 83 - 94. El-Sebaii AA, Trabea AA (2005). Estimation of Global Solar Radiation computer software program. It was observed that equa- on Horizontal Surfaces Over Egypt, Egypt. J. Solids 28: 166. tions (9), (10) and (11) have the highest values of correla- Exell RHB (2000). The intensity of solar radiation. King Mongkut’s tion coefficients and coefficient of determinations for Ca- University of Technology Press, Thornburi Pp. 549-554. labar, Port Harcourt and Enugu respectively which gives Fagbenle RO (1990). Estimation of total solar radiation in Nigeria using Meteorological data. Nig. J. Renewable Energy 1: 1-10. good results when considering statistical indicators, that Folayan CO (1988). Estimation on global solar radiation bound for some is, MBE, RMSE and MPE. The regression equations for Nigerian cities. Nig. J. Solar Energy 3: 3-10. Calabar, Port Harcourt and Enugu with the smallest va- Halouani M, Nguyen CT, VoNgoc D (1993). Calculation of monthly lues of MBE, RMSE and MPE are average global solar radiation on horizontal surfaces using Daily Hours of Bright Sunshine, Solar energy 50: 247-255. Ibrahim SMA (1985). Predicted and measured global solar radiation in 2 Egypt. Solar Energy 35(2): 185 -188. ΗΡ n ≈ n ’ Igbal M (1983). An introduction to solar radiation. Academy press. New = −0.499 + 4.294 − 4.769∆ ÷ , york. pp. 6-51. ΗΟ Ν «Ν ◊ Iqbal M (1979). A study of Canadian ucu diffuse and total solar radiation data, Solar Energy 22 (1): 81. Ojosu JO (1984). Solar Radiation Maps of Nigeria. Nig. J. Solar Energy 2 8: 370 – 384. ΗΡ n ≈ n ’ = 0.177 +1.048 −1.070∆ ÷ and Prescott JA (1940). Evaporation from a water surface in relation to solar radiation, Tran. R. Soc. S. Austr. 64: 114 -118. Ο Η Ν «Ν ◊ Sambo AS (1985). Solar radiation in kano: a correlation with 2 meteorological data. Nig. J. solar Energy 1: 59. ΗΡ n ≈ n ’ Shears RD, Flocchini JL (1981). Technical note on correlation of total, = 0.177 +1.048 −1.070∆ ÷ respectively. diffuse and direct solar radiation with percentage of possible ΗΟ Ν «Ν ◊ sunshine for Davis, california. Solar Energy 27(4): 357-360.

The equation could be employed in estimating global solar radiation of location that has the same geographical location information as Calabar, Port Harcourt and

Enugu.

ACKNOWLEDGEMENT

The authors wish to express their gratitude to the management and staff of the Nigerian Meteorological

Agency, Oshodi, Lagos, State for their kindness in supplying the data for sunshine hours for this work and the renewable energy for rural industrialization and development in Nigeria for making the global solar radiation data available.

REFERENCES

Adebiyi GA (1988). An empirical correlation of solar radiation data for Nigeria. The Nigerian Engineer 32(2): 50 - 57.